OpenAI has announced the introduction of a memory feature for ChatGPT. This new capability allows the AI to remember user inputs and utilize this information in future interactions, aiming to make conversations more coherent, personalized, and useful over time.

The memory feature in ChatGPT operates in two distinct modes — manual and automatic. In the manual mode, users can explicitly tell ChatGPT to remember specific details. This could range from personal preferences to professional details, enabling the AI to tailor its responses more accurately to the user’s context. The automatic mode, on the other hand, does not require any direct instruction from the user. Instead, it learns and remembers based on the ongoing interaction, gradually building a reservoir of knowledge about the user.

Despite the potential benefits of a more personalized AI experience, the introduction of memory capabilities raises significant privacy concerns. Recognizing this, OpenAI has implemented measures to give users control over this feature. Through the settings menu, individuals can manage their memories with ChatGPT, including the ability to view, delete, or even disable the memory feature entirely. This ensures that users can enjoy the benefits of personalized interactions without compromising their privacy.

The organization assures that ChatGPT is designed to avoid storing sensitive personal information and that users have the final say in what information is remembered. Moreover, OpenAI has stated that while the content provided to ChatGPT, including memories, may be used to improve the model for all users, individuals have the option to opt-out through their data control settings.

The memory feature is initially being made available to a select group of subscribers, including those on the Plus plan and free users. This phased approach allows OpenAI to refine the feature, address any issues, and ensure that it meets the high standards expected by its user base before a broader rollout.